Updated on Sep 30 2016

These are preliminary results obtained by applying a Resource Selection Function (RSF) for Caribou (forest-dweling ecotype) to a LANDIS-II simulation ensemble in the Lac St-Jean area.

Description of simulation ensemble

 

Projected climate

Factor Levels Number of levels
Climate change scenario    RCP 2.6, RCP 4.5, RCP 8.5 3
Fire regime baseline, projection1 2
Harvesting levels 50%, 100% 2

1Fire regimes projections are based on the climate change scenario being simulated, i.e. corresponding RCP.

  • 12 treatments with projected climate (3 scenarios x 2 fire regimes x 2 harvesting levels)

 

Baseline climate

Factor Levels Number of levels
Climate change scenario    baseline       1
Fire regime baseline 1
Harvesting levels 50%, 100% 2
  • 2 treatments with baseline climate (1 scenario x 1 fire regime x 2 harvesting levels)

 

Summary

Each of the 14 treatments was replicated 5 times.

  • Total (expected): 70 simulations (3 simulations failed)
  • Total (final): 67 simulations

Thus, the entire simulations ensemble was planned to contain 70 simulations. However, three of those failed which caused the number of simulations to be reduced to 67.

*Climate change scenarios, fire regimes, and harvesting levels define each simulated treatment.

  • Each treatments are replicated 5 times (with a few exceptions).

Projected climate (RCPs): 3 baseline simulations:

  • A total of 70 simulations were planned (3 of which failed)
  • A total of 67 simulations make up the simulation ensemble.

More details about the RSF itself and its implementation can be found here (to be completed soon).

Simulation ensemble description

If you prefer visualizing the following results with static figures instead of animations, click on the links below each one of them (or clone the entire repo).

Results (preliminary)

Mean probability of occurrence (averaged over entire landscape)

The harvesting levels, relative to recent year levels, are the most important determinant of Caribou’s future probability of occurrence when averaged over the entire landscape. Caribou’s future probability of occurrence is inversely proportional to harvesting level.

Harvesting levels are followed by climate change scenarios and fire regimes, which both also have subtantial impacts on Caribou’s future probability of occurrence. The detrimental effects of climate change and fire regime increase with the intensity of the radiative forcing.

Future probability of occurrence remains stable only under the baseline climate change scenario, when harvesting levels are set to 50% compared to recent year levels.

Static figures - Entire landscape (averages)

High quality habitats (averaged over entire landscape)

The abundance of high quality habitats (HQH), defined using thresholds corresponding to the top 10% and top 25% habitats in the initial landscape, is mostly affected by fire regime, closely followed by harvesting levels. (Each curves indicate the proportion of the landscape that contains habitats with RS probability greater than those associated with the top 10th and 25th percentiles in the initial landscape.)

Static figures - Entire landscape (High quality habitats)

The response of Caribou future resource selection to the various sources of uncertainty will be heterogenous throughout the landscape.

Mean probability of occurrence (averaged by ecodistrict)

While most of the ecodistricts follow the same trends as those described for the entire landscapes, the northermost ecodistricts (432, 435, 436) appear to be most sensitive to changes in fire regimes.

Static figures - By ecodistrict (averages)